82 research outputs found
A novel Bayesian strategy for the identification of spatially-varying parameters and model validation in inverse problems: an application to elastography
No full textThe present paper proposes a novel Bayesian, computational strategy in the context of model-based inverse problems in elastostatics. On one hand we attempt to provide probabilistic estimates of the material properties and their spatial variability that account for the various sources of uncertainty. On the other hand we attempt to address the question of model fidelity in relation to the experimental reality and particularly in the context of the material constitutive law adopted. This is especially important in biomedical settings when the inferred material properties will be used to make decisions/diagnoses. We propose an expanded parametrization that enables the quantification of model discrepancies in addition to the constitutive parameters. We propose scalable computational strategies for carrying out inference and learning tasks and demonstrate their effectiveness in numerical examples with noiseless and noisy synthetic data.<br/
Hertzian Fields: Exploring WiFi microwave signals as a spatial and embodied sensing medium for art
No full textThesis (Ph.D.)--University of Washington, 2023This dissertation is centered around a series of three artworks (Hertzian Fields) that explore WiFi as a spatial and embodied sensing medium. These works use a new sensing technique developed by the author that leverages the interference of the human body on WiFi signals to create highly responsive live performance and interactive systems. Hertzian Field #1 (2014) is an augmented reality immersive environment using sound to explore the materiality of WiFi communication through its interaction with space and the human body. Hertzian Field #2 (2016) is a 20'-25' augmented reality immersive performance for solo performer, WiFi fields, computers and surround sound that conjures a phenomenology of the hertzian medium explored through sound and movement. The Water Within (Hertzian Field #3 and #3.1) is a reactive wet sauna: an intimate multi-sensory environment of complete immersion, combining WiFi sensing fields, machine listening software, embedded 3D sound, hot steam, and architectural design. Steered by the flows and variable densities of water molecules traced in steam and bodies by (ab)using WiFi, it creates a regenerative post-relational experience that celebrates interference, signal-loss, and disconnecting. The piece exists in two iterations and formats: an interactive installation (2016) and a composed interactive experience (2018).
The dissertation describes the author's conceptual and technical approach in using WiFi microwave signals as an artistic medium. It also examines the background, context, ideas and research processes that led to the creation of these works. In doing so, it lays the foundation for developing a better and deeper understanding of microwaves and WiFi signals, investigates their artistic potential, and discusses related approaches by other artists.
Chapter One (Introduction: The hertzian medium) introduces core ideas and concepts regarding the medium. This includes: a discussion on the impact of wirelessness in contemporary living and how it has transformed our interactions with and understanding of the world; an overview of the physics of electromagnetism and the electromagnetic spectrum; and an investigation of the hertzian (i.e. radio and microwaves) as a multilayered medium consisting of seven interconnected layers: physics, science, imagination, engineering, use, impact, regulation.
Chapter Two (The birth of a medium: Energy becomes technology) introduces a media archaeological approach as a method for grasping what the medium affords, and how our imagination of what we can use it for has developed over time. It presents an overview of key developments in hertzian science, imagined and realized applications, and their impact. This chapter focuses primarily on the early years around Heinrich Hertz’s discovery of electromagnetism, looking at the birth of wireless technologies relevant to the Hertzian Field series: communication, broadcast, hacking and electronic warfare, navigation, meteorology, radio astronomy, and radar, before closing with a section on the development of WiFi.
Chapter Three (Radar and Direction-Finding in sonic art and beyond) surveys musical instruments and artworks based on spatial and/or embodied uses of the hertzian as a sensing medium. The emphasis is on sound-centric practices and specific technologies that have been used to this extent: from capacitative / electric field sensing, to musical instruments utilizing direction-finding principles, to spatial uses of broadcast radio, to doppler radar systems. Instruments discussed include: Theremin and Terpsitone; Pupitre d'Espace; Radio Baton; Marimba Lumina. Artworks by the following artists are examined: Max Neuhaus; Edwin van der Heide; Christina Kubisch; John Cage; Philippa Cullen; Liz Phillips; Sonia Cillari; Tetsuo Kogawa; Anna Friz; Edward Ihnatowicz; Steve Mann; Joe Paradiso / MIT Lab; Arthur Elsenaar; Godfried-Willem Raes.
Chapter Four (First hertzian explorations: From the network to the body, from WiFi to Radar) turns to the author's own work. It presents the first phase (2010-14) of his research trajectory on the hertzian medium, and introduces three projects in which he explored WiFi and broadcast radio.
Chapter Five (Ubiquitous sensing with radio waves and microwaves) dives into the technological context influencing the author's research. It introduces the field of Ubiquitous Sensing and discusses relevant localization and device-free sensing techniques, concluding with a discussion on the physics and biological factors involved so as to comprehend how and why such techniques work.
Chapter Six (Wireless Information Retrieval: Sensing with WiFi signals) presents the device-free WiFi-sensing technique that the author developed for the Hertzian Field series. Combining elements from Ubiquitous Sensing and Music Information Retrieval, this technique performs multi-layered feature extraction on the Received Signal Strength Indication (RSSI) of WiFi Beacon frames to deduce a variety of information related to the movement of uninstrumented bodies, and to changes in environmental factors (e.g. humidity).
Chapter Seven (Composing Hertzian Fields) discusses strategies for creating works with this technique, and examines the three works of the Hertzian Field series in detail. It finally touches on ideas for future work by the author
On the Achievable Rates of OFDM with Common Phase Error Compensation in Phase Noise Channels
No full textThis paper considers the problem of analytically assessing the maximum achievable rates (capacity) of orthogonal frequency division multiplexing (OFDM) transmissions when the receiver, for complexity reasons, only accounts for the common phase error (CPE) effect due to phase noise impairments with inter-carrier interference (ICI) treated as noise. By recognizing that the functional form of the CPE with respect to the phase noise realization is actually a free design parameter, determination of the capacity is posed as a functional optimization problem with respect to the, so-called, CPE function. A simple lower bound of the capacity is obtained, revealing the performance degradation due to the unknown CPE at the receiver, as well as the suboptimal performance achieved in severe phase noise conditions by the conventional CPE function that is routinely employed in previous works. The existence of an optimal number of subcarriers that balances the effects of the (unknown) CPE and ICI is highlighted and critical system design/operation issues, such as selection of the CPE function and effect of unknown channel on the achievable rate, are discussed. The analysis in this paper can be employed for determining the suitability of OFDM in phase noise channels and provides a tractable utility function for resource allocation purposes.Manuscript received October 7, 2016; revised February 10, 2017 and May 5, 2017; accepted May 6, 2017. Date of publication May 17, 2017; date of current version August 14, 2017. This research work was made possible by grant number NPRP 6-1326-2-532 from the Qatar National Research Fund, QNRF (a member of the Qatar Foundation, QF). The statements made herein are the sole responsibility of the authors. The associate editor coordinating the review of this paper and approving it for publication was G. Colavolpe. (Corresponding author: Stelios Stefanatos.) S. Stefanatos is with the Department of Digital Systems, University of Piraeus, 18534 Piraeus, Greece (e-mail: [email protected]).Scopu
A Greek – Makriyannis
No full textTitle: Ἕνας Ἕλληνας - ὁ Μακρυγιάννης (Α Greek – Makriyannis) Originally published: As a leaflet in Alexandria, 1943 Language: Greek The excerpts used are from Γεωργίου Σεφέρη, Δοκιμές τ. Α,’ (Athens: Ἴκαρος, 1974), pp. 236–237, 240–241, 257–259, 261–262. About the author George (Georgios) Seferis [1900, Smyrna (Tur. Izmir, present-day Turkey) – 1971, Athens]: poet, literary critic and diplomat. His father Stelios Seferiadis was a famous lawyer as well as a translator and poet. The family mili..
A Greek – Makriyannis
No full textTitle: Ἕνας Ἕλληνας - ὁ Μακρυγιάννης (Α Greek – Makriyannis) Originally published: As a leaflet in Alexandria, 1943 Language: Greek The excerpts used are from Γεωργίου Σεφέρη, Δοκιμές τ. Α,’ (Athens: Ἴκαρος, 1974), pp. 236–237, 240–241, 257–259, 261–262. About the author George (Georgios) Seferis [1900, Smyrna (Tur. Izmir, present-day Turkey) – 1971, Athens]: poet, literary critic and diplomat. His father Stelios Seferiadis was a famous lawyer as well as a translator and poet. The family mili..
BECLR: Batch Enhanced Contrastive Unsupervised Few-Shot Learning
No full textThere exists a fundamental gap between human and artificial intelligence. Deep learning models are exceedingly data hungry for learning even the simplest of tasks, whereas humans can easily adapt to new tasks with just a handful of samples. Unsupervised few-shot learning (U-FSL) aspires to bridge this gap, without relying on costly annotations. Inspired by the efficiency of contrastive representation learning, we propose a novel batch enhanced contrastive U-FSL pretraining methodology (coined as BECLR) to infuse instance- and class-level insightswithin a contrastive framework. To enable the sampling of meaningful positives, we introduce an innovative dynamic clustered memory module (DyCE), which maintains highly-separable latent space partitions, through iterative equipartitioned updates. We also propose an effective, optimal transport (OT)-based feature alignment strategy (OpTA), to address sample bias in the U-FSL inference stage and further boost the end-to-end performance of BECLR in low-shot settings. Our extensive experimental evaluation corroborates the efficacy of our design choicesin BECLR, which sets a new state-of-the-art on the most widely adopted U-FSL benchmarks miniImageNet and tieredImageNet (offering up to 14% and 12% improvements, respectively), as well as on challenging cross-domain scenarios.Computer Scienc
Anamnesis of tomorrow: Rediscovering the raison d’être of the National Archives of the Netherlands in the age of digital transformation.
No full textAn archive is a repository of documents, usually of historical importance. Therefore, one could say that the National Archive of the Netherlands exists as a result of the documents it holds. In fact, its contents include: 137 km of documents, 15 million photographs, 300,000 historical maps and drawings, and 800 terabyte of digital files. However, as our everyday life becomes more and more digital, the National Archive faces a distinctive upcoming change; transitioning from document-centredness to data-centredness. Within this framework, this project aims to answer the following question: Do archival buildings have a future in the Digital Age? Can archives exist without documents? Or are they destined for obsolescence? Through a qualitative research in the cultural value and societal role of archives the project seeks answers to the above questions. In continuation, new questions are brought up which aim to direct the project in an architectural conclusion. Whether the Archive remains as we know it or changes, one thing is sure; the purpose of the existing building has to be redefined. In this pursuit, several ingredients are put together which relate to social, contextual, structural, architectural and sustainability issues. The project considers not only the re-design of the National Archive building but the wider re-design of the site, as an important piece of urban fabric to the city of the Hague.All things considered, the National Archive is to be visualised as a more public place and a piece of public infrastructure that responds to a multiplicity of challenges.Architecture, Urbanism and Building Sciences | Building Technolog
Iron and arsenic removal in triple- and single-bed filters: The effect of pH, filtration velocity and filtrate recirculation
No full textArsenic in groundwater can constitute a persistent nuisance for water treatment facilities when it exceeds the admissible limit of 10 μg/L. Recently, a stricter limit has been set as a new challenging target by many companies in the Netherlands, which is below 1 μg/L. However, most of the groundwater treatment plants have been conventionally designed solely for the removal of the most common undesirable groundwater constituents, namely iron, manganese and ammonium. The current research aimed at the investigation of the operational conditions facilitating As removal in biological rapid filters simultaneously with the required Fe removal. This improved As retention should be correlated with an extended length where its adsorption takes place, thus with the deeper Fe penetration inside the bed. Therefore, this was attempted in this research. The process water used throughout the experimental tests contained As(III) and Fe(II) in order to simulate a typical anoxic groundwater quality. For the most part, a triple-layer filter bed was used, consisted of anthracite, sand and garnet. The different settings under examination involved a range of pH values (7.8, 7.1 and 6.4), two filtration velocities (2.5 m/h and 5 m/h) as well as the recirculation of the filtrate back to the feed stream. Finally, the multimedia bed was compared with a single-layer sand filter. The results of the conducted pilot column filter experiments revealed that high pH values were accompanied with high oxidation rates and thus with the creation of Fe flocks, already in the supernatant water. Due to this, at pH 7.8 and 7.1 lower Fe concentrations were detected in the effluent, denoting a higher Fe retention as compared to pH 6.4. Interestingly enough, a relatively deep Fe penetration was observed for every pH value tested. Regarding As removal, it was evidently favored by high pH values owing to the oxidation-floc formation removal mechanism of Fe (homogeneous reaction), which prevailed under those conditions. On the other hand, at pH 6.4 the adsorption-oxidation mechanism was predominant (heterogeneous reaction), which obstructed the AsOB activity. As far as the tested filtration velocities is concerned, they didn’t seem to significantly impact both Fe spread over the bed as well as As removal at the higher pH levels. Nonetheless, this was not the case for pH 6.4, in which the slow flow rate enabled the generation of more and larger Fe flakes (due to sufficient residence time in the supernatant water), which were then retained in the top part of the bed. The high flow rate on the other hand allowed Fe to reach deeper in the filter. Surprisingly, As removal seemed to be improved at 2.5 m/h, despite the Fe flocks accumulation in the upper layers. Possibly, the short experimental times not allowing equilibrium to be reached could comprise a reasonable explanation of this unexpected result. Furthermore, the filtrate recirculation stream didn’t seem to positively influence As removal. The induced dilution effect resulted in a relatively large dispersion of Fe inside the filter bed, however the essentially halved incoming Fe concentration was not sufficient to adsorb the oxidized As(V). Finally, the comparison between the multimedia bed with the single-layer filter reveals a considerably wider Fe dispersion over the bed height in the former case, which in its turn promotes a more efficient As removal. The overall conclusion of the current study is that triple-layer bed filters facilitate a more gradual Fe removal and its deeper penetration in the bed as compared to single-layer filters. This fact stimulates As removal and additionally allows for longer filter run times. Moreover, heterogeneous Fe removal seems to obstruct As oxidation by AsOB and therefore homogeneous reaction can be considered as more favorable in terms of As removal. This specific removal mechanism becomes predominant at pH levels above 7, when sufficient oxygen is available. Lastly, the operational setting of filtrate recirculation back to the filter inlet, displays a negative impact regarding As removal.Civil Engineerin
Causal Inference in DotA 2 when estimated through randomized data
No full textStrategy games could be considered as an amazing playground for using Causal inference methods. The complex nature of the data and the built-in randomization help with testing causal inference in a scenario where in reality it would be hard and expensive. Randomized data in coherence with causal inference is well documented and tested, but not regarding the strategy game of interest DotA 2. To evaluate the quality of causal inference using randomized data for predictions in the game, the average causal effect estimand is used. The calculation of the average causal effect of certain events between different intervals and their comparison in addition to the calculation of the statistical Independence between variables of concern comprise the bulk of the research. The calculations allow for logical deductions and statistical correlations between values to reach a conclusion. The final verdict being that causal inference with randomized data is helpful for predicting events in DotA 2 but the amount of data and existing complex biases can be deceiving and can heavily influence the results.https://github.com/stelios34S/Causal-inference-in-DotA-2-when-estimated-through-randomized-data.git Source Code RepositoryCSE3000 Research ProjectComputer Science and Engineerin
Mitigation of Silica Scaling by Closed-Circuit Reverse Osmosis
No full textReverse osmosis (RO) is considered the most reliable and cost-effective membrane desalination technologyworldwide. However, it suffers significant performance limitations due to mainly inorganic foulinggenerated in the highly concentrated brine. Especially, scaling caused by silica and silicates depositionsresults in irreversible damages with considerable economic implications. Recently, a different ROconfiguration, termed as closed-circuit reverse osmosis (CCRO), has been claimed to exhibit substantialbenefits over conventional RO in terms of both energy savings as well as higher scaling resilience.CCRO is operated in batches, during which the generated brine is continuously recycled inside theclosed loop until a desired recovery has been accomplished, after which the brine is released and replacedby fresh feed. Regarding CCRO scaling resistance superiority, an experimental-based proofis missing from the relevant literature. The current thesis was realized in collaboration with LenntechB.V., aiming at investigating the intrinsic propensity of CCRO to withstand and delay silica scaling. Tothat end, a campaign of filtration tests was carried out by means of a single-module CCRO pilot setup,during which two scaling indicators were periodically monitored. The used indicators were the masstransfer coefficient (MTC) and the applied feed pressure (Pfeed). Prior to the filtration trials, preliminarybatch tests, of 4-hour duration each, were carried out in order to simulate and more thoroughly examinethe circulated brine conditions. Various synthetic brines were prepared and silica polymerization wasmonitored. The effects of silica supersaturation level, pH and hardness ions were investigated. Of greatimportance was whether silica existed in its monomeric or polymeric form, since this greatly impactsthe scaling occurrence probability. Batch tests results revealed that at high pH conditions (pH>10)monomeric silica concentration remained unchanged in pure silica solutions (even at high supersaturationlevels), owing to the great silica solubility level. Nevertheless, when Mg2+ and/or Ca2+ werepresent in the solution, the quantity of silicic acid rapidly reduced. This was the result of the instantaneousformation of metal-silicate precipitates. Batch tests at pH 7 were also performed. In that case,monomeric silica concentration in pure silica solutions remained constant up to initial concentrations ofabout 450 mg/L SiO2 for the examined 4-hour duration. However, at higher SiO2 concentrations, suchas at 750 mg/L, rapid polymerization occurred. When hardness cations were included in the neutral pHsolutions, they showed an accelerating effect on silica polymerization process, but they did not reactwith either monomeric or polymeric silica. This effect relates to the suppression of the silica colloidsdiffuse double layer by the hardness cations, which subsequently facilitates colloids agglomeration.Regarding the CCRO filtration tests, they were conducted in sequences with duration of 20 or 40 min,which in its turn determined the achieved sequence recovery. For most of the carried out sequencesthe initial feed composition was: 120 mg/L SiO2 and 24 mg/L Mg2+. Only the final 5 out of the total40 sequences were realized in the absence of magnesium in the feed solution. All the filtration runswere performed at pH 7, at ambient temperature and at constant flux 15 L/m2h. The outcome was ascaling-free desalination process for a total cumulative operational period of approximately 11 hours,during which recoveries as high as 90.9% were reached, whereas severe scaling took place only afterabout 14 hours of total operation. The obtained results were contrasted with filtration tests results ofconventional RO received from literature resources and in that way the higher efficiency of CCRO towithstand and delay silica scaling was proved. Additionally, through silica mass balance calculations itwas shown that during all filtration tests significant silica polymerization took place. Also, cations analysisbe means of IC excluded the participation of Mg2+ ions in the formed scale layer. It was concludedthat the scale development was the result of an initial attachment of silica colloids to the membranesurface followed by monomeric units adsorption onto them. Finally, a simple customized method forthe prediction of silica scaling potential in CCRO operations based on batch tests was proposed.Civil Engineering | Environmental Engineerin
- …
